AI could help ‘diagnose dementia in a day’

AdobeStock_226212015-scaled-2

By Ed Pinches | Tuesday 10 August 2021

Researchers at Cambridge University and Addenbrooke’s Hospital in Cambridge have developed an approach using artificial intelligence that can diagnose dementia from a single brain scan, as well as providing doctors with information about that person’s prognosis.

Prof Zoe Kourtzi, a researcher at Cambridge University and a fellow of the national centre for AI and data science, The Alan Turing Institute, developed the algorithm. Prof Kourtzi is also Scientific Director of Early Detection of Neurodegenerative diseases (EDoN), a global research project Alzheimer’s Research UK are spearheading and combines digital data from wearables to predict who may go on to develop dementia years before symptoms show.

The Cambridge University team will now work with doctors at Addenbrooke’s Hospital in Cambridge to test the AI approach in a real-world setting, with patients coming into the memory clinic with concerns about dementia.

What our expert had to say?

Dr Laura Phipps at Alzheimer’s Research UK, said:
“Artificial intelligence systems like this are currently being applied to many aspects of healthcare, drawing on the insights from huge datasets to help doctors make more informed decisions about diagnosis, treatment and care.

“To diagnose dementia today, doctors need to rely on the interpretation of brain scans and cognitive tests, often over a period of time. Machine learning models such as those being developed by Prof Kourtzi could give doctors greater confidence in interpreting scans, leading to a more accurate diagnosis for patients.

“We hope that in future, such approaches will not only improve how current diagnostic techniques are implemented but open the door to revolutionary new approaches to detect the diseases that cause dementia much earlier. This would have a huge impact on people with dementia and their families.

“Artificial intelligence systems also make it possible to combine many different data types to accurately predict who might have a particular disease. Prof Kourtzi is leading an Alzheimer’s Research UK funded project called EDoN, combining digital data from apps and wearables including sleep, cognition, fine motor skills and brain activity to predict diseases like Alzheimer’s 15-20 years before we can today.”

How difficult it can currently be to get an accurate diagnosis

Denise Wallin’s experience with her mum, Barbara, shows how difficult it can be to get an accurate diagnosis of dementia.

It took four years for Barbara to be diagnosed with dementia, and she was first wrongly diagnosed with vascular dementia before this was changed to frontotemporal dementia (FTD). She was aged 60 when she was diagnosed in 2012 and died in 2018.

Denise Wallin said:

“My mum was first diagnosed with depression, but I was convinced it was something more and thought it might be dementia.

“She did the memory tests, but she sailed through them, as FTD affects behaviour more than memory in the early stages.

“My dad, my brother and me would take it in turns to take her to her GP and we had to nag and nag him to send her for a scan.

“Eventually she had a brain scan and she was diagnosed with dementia, however she was originally diagnosed with vascular dementia.

“I read up about dementia and found the information about the different forms of dementia on the Alzheimer’s Research UK website really helpful. As I learned more I felt the vascular dementia diagnosis just didn’t add up.

“She was referred to a hospital consultant and her diagnosis was changed to FTD. It was a relief to finally get the diagnosis, although it was very sad. There aren’t any treatments specifically for FTD, but an early, accurate diagnosis would have avoided a lot of stress and helped us prepare for what was to come.”

Want to know more?

You can read about Alzheimer’s Research UK’s Early Detection of Neurodegnerative diseases here: edon-initiative.org/

Tags:

By submitting a comment you agree to our comments policy.
Please do not post any personal information about yourself or anyone else, especially any health data or other sensitive data. If you do submit sensitive data, you consent to us handling it in line with our comments policy.

Leave a Comment





About the author

Ed Pinches